Discussion of Effective Health

I am sure about the diminishing returns of armor, crit, resist, and block rate/block amount. Please look at the link I had in my second paragraph in the previous post. That is where I display some data for LVL 40. (Before LVL 45 was released) The short version is that to get from 0% damage reduction to 40% damage reduction might only take 1500 armor (rough number), but doubling your armor to 3000 only gets you about 60% damage reduction (not 80% which coincidentally is the maximum reduction imposed by the game programing). That is why after a certain point it becomes more beneficial to get HP and not armor. It is also why people almost never use cyanites and prefer to use their slots for Resist and HP.

The purpose of the DPS formula I gave is not designed to take into account all factors (cast times that are unaffected by attack speed, Damage over time attacks, Higher essence use of low damage/high speed builds... etc) It is simply for a direct compareison of how much damage on average a particular build can pump out. It lets you compare changes in stats. For example, if you had a change that would gain 4.2% attack speed, lose 150 crit, gain 3% crit damage and lost 10 damage, would you know if this would be an improvement or if you would end up losing damage total? By plugging in your stats before and after these changes, you can compare the DPS numbers to see if you would have a net gain or net loss of DPS taking into account all offensive stats. This is already normalized to be in Damage/Second so dividing by attack speed is irreverent and counter productive as it would cancel out one of the most important components of DPS... the speed! Furthermor, you say one can easily boost their attack speed... I disagree. For endgame players with nice gems (Radiant and Flawless), any significant change in speed often comes at the expense of damage, crit, or crit damage and when you need have those changes, you can use the equation above to see just how significant those changes are.

A real example from my character. I considered the Dragonache set. I would have gained speed, lost damage, gained crit, and lost crit damage. When I plugged all the numbers in, I actually lost about an average of 50 DPS so I decided against getting the set. That is a perfect example of speed going up but other stats going down enough to make it not worth it.

 

i calculated this in excel to demonstrate what I mean:



Let's look at the "DR increase per 500 AV" column. As you said, every 500 additional armor points yield less and less additional DR% (20,67 / 13,58 / 9,61 and so on).
That's what you call diminishing return, which is wrong if we look at the effective health (EH) increase.
Let's look at my 5k life example below. Our effective health increases linearly by 1302,76 life for every 500 armor value points spent.

So one may decide whether to spend 500 armor points or 1303 life to get the same effective health. Of course this isn't correct anymore if you go beyond 80% (but well who does that)

 

What equation are you using for "Effective Health" and how are you deriving it?

 

The equation is hp/(1-dmg reduction), but what Aemon said would make sense only if the magical dmg didn't exist.

 

Aemon is right; when magical damage is concerned magic resistance values have to be used instead of armor value. In both cases there is no diminishing returns : the higher armor/magic resist, the more damage you can take.

 

Ok, I see the equation and get where it is coming from, but it makes me agree that because of elemental damage effective HP can not solely be based on armor. I'd say it depends on the the kind of attack. I remember on the old forums someone doing some work and looking at elemental damage. If I remember correctly, Lighting goes through armor and is totally dependent on resistance where as fire and ice were partially dependent on resist and armor (50/50?). Nothing was explored when it came to poison or andermagic. That would mean that a character could have multiple Effective HP values depending on the kind of attacks they are receiving. In that case, a DK fighting a fire build mage would have a different Effective HP than if he were fighting a lighting build mage, and an even different one if he were fighting another tank or a ranger. Looks like I have a new math problem to consider.

One final though on HP vs Armor... the HP regeneration of Supernatural Regeneration, Furious Battle Cry, and Dragon Hide is based off of total HP, not Effective HP. You can have 5k armor and only 5k HP and the calculations might say you have better effective HP than another character with 2k armor and 10k HP (didn't do the math here, I'm in a rush, just estimating), but the DK with 10k HP will heal twice as much from the three Regeneration skills that are a percentage of your total HP when compared to the heavily armored one.

On a side note, can a mod move this discussion of HP over to a new thread and leave the Damage discussion here? Thanks.

 

Indeed the are just 5 different "effective HP" (I'd rather call it "endurance") values per character, there are no maths to do here.
It's not elements that are dependent on armor/resist, but some skills that are partially physical AND elemental : Ranger's mighty shot, SW's fireball... (I might be the guy from the old forum). These skills partially hit 2 of your 5 effective HP bars.
HP regeneration has nothing to do in all this; you simply regenerate x% of your effective HP (if I remember well, effective HP's formula is something like "HP * (1918 + Armor)" at level 45).

 

Ofc that was just an example for armor to show that there's no "diminishing returns"...
To calculate your effective health correctly, you need to consider resist reduction and block chance/amount as well.

/sign

 

My point was:

• fighting a warrior 500 armor = 1300 hp
• fighting a ranger 500 armor < 1300 hp
• fighting a mage 500 armor <<< 1300 hp